声脉冲辐射力成像是一种以超声为基础的弹性成像工具，可对肝纤维化进行十分准确的无创评估。其他可供选择的弹性成像检查如Fibroscan®，虽临床应用更加广泛，但其对体质指数(BMI)较高患者的图像获取率及图像可信度较低。因此，声脉冲辐射力成像更适用于这一不断增长的人群。遗憾的是，目前还没有充分的关于肥胖对ARFI成像影响的相关数据来支持ARFI对高BMI患者的临床应用。我们把皮肤到肝包膜的深度作为中心性肥胖的替代性指标，通过测量SLD，来检测中心性肥胖对ARFI成像是否有影响。研究结果将发布在AASLD2014上。
研究方法
我们的试验纳入了772位经ARFI成像的病因不同的慢性肝病患者。所有的患者都经由2位或2位以上不知情操作者进行测试，共产生了1902份测量结果。皮肤表面到肝脏包膜的距离在ARFI画面截图上进行独立测量。有55位患者，在进行了6个月的ARFI同时，也进行了肝脏活检，使这项亚组研究具备了直接的组织学相关性。
研究结果
我们的队列研究反映出了很强的人口变化趋势，其中有68%的患者超重(BMI>25)，42%的患者肥胖(BMI>30)。皮肤-肝脏距离(SLD)紧密服从体质指数BMI (皮尔森相关系数：0.818, P值<0.001),SLD大于2.5cm与BMI大于30相对应。SLD与测量离差密切相关(即四分位数间距, p=0.0001)，在SLD<2cm、SLD介于2- 2.5cm、 SLD >2.5cm的患者中，测量结果的四分位数间距大于0.3的比例分别为20.3%, 43.9% 和 77.1%。
SLD也与操作者之间测量结果的一致性相关(p=0.0001)，在SLD均值分别为小于2cm, 介于2-2.5cm 和 大于2.5cm的患者中，操作者测量结果不一致的概率分别为4.83% (95%CI 3.03-7.55%)、16.30% (95%CI 12.63-20.77%) 、32% (95%CI 14.50-25.27) 。此外，随着SLD的增加，ARFI的准确性会降低(p=0.009)。SLD均值分别小于2cm, 介于2-2.5cm 和 大于2.5cm时，ARFI成像结果与组织活检结果一致的概率分别为87.18%(95%CI 75.62-94.87), 76.92% (95%CI 63.72-86.42) 和48.15% (95%CI 28.67-68.05)。尽管BMI对这三种指标有着相似的影响，但回归分析发现SLD能够更好的预测ARFI结果的可信度。
结论
以SLD为指标的向心性肥胖与ARFI成像效果有着明显的相关性。作为成像可靠性指标，SLD比BMI更具优越性。大于2.5cm的SLD与较大的四分位数间距、增高的操作人员测量结果不一致率及较低的成像总准确率相关。因此，在我们向SLD大于2.5cm的患者解释ARFI结果时，应注意到这些。
英文原文
Central obesity negatively impacts the performance of Acoustic Radiation Force Impulse Imaging (ARFI) in the assessment of liver fibrosis.（摘要号：LB-37）
Abstract Body: Background: ARFI is an ultrasound-based elastography tool, which has shown excellent accuracy in the non-invasive assessment of liver fibrosis. Whilst alternative elastography tests such as Fibroscan® are more widely available, these have shown lower acquisition rates and reliability in patients with a high body mass index (BMI). ARFI is therefore often favored in this increasingly large population subset. Unfortunately, there is currently insufficient data regarding obesity’s influence on ARFI performance to support this practice.
Aim: To determine whether central obesity has an impact on ARFI performance, by using skin-to-liver capsule depth (SLD) as a surrogate marker for truncal adiposity.
Method: 772 patients with mixed etiology chronic liver disease underwent ARFI within our institution. All patients were tested by two or more blinded operators, resulting in a total of 1902 measurement sets. The distance between the skin surface and liver capsule was independently measured from ARFI screenshots. Fifty-five patients had also undergone liver biopsy within six months of ARFI, allowing direct histopathologic correlation in this small study subset.
Results: Our cohort mirrored greater population trends, with 68% of patients being overweight (BMI>25) and 42% obese (BMI>30). Skin-liver distance (SLD) closely followed BMI (Pearson’s r=0.818, p<0.001), with a SLD>2.5cm approximating a BMI>30. SLD correlated strongly with measurement dispersion (i.e. IQR, p=0.0001), with the proportion of measurement sets having an IQR>0.3 being 20.3%, 43.9% and 77.1% in patients with SLD<2cm, 2- 2.5cm & >2.5cm respectively.
SLD also correlated with inter-operator concordance (p=0.0001), with discordance rates being 4.83% (95%CI 3.03-7.55%), 16.30% (95%CI 12.63-20.77%) and 19.32% (95%CI 14.50-25.27) in patients with a mean SLD of <2cm, 2-2.5cm and >2.5cm. Furthermore, ARFI showed lower overall accuracy with increasing SLD (p=0.009). Rates of concordance with biopsy were 87.18% (95%CI 75.62-94.87), 76.92% (95%CI 63.72-86.42) and 48.15% (95%CI 28.67-68.05) when SLD was <2cm, 2-2.5cm and >2.5cm respectively. Whilst BMI had a similar influence on all three performance measures, regression analyses found SLD to be superior in the prediction of ARFI reliability.
Conclusion: Central obesity as indicated by SLD, correlates strongly with ARFI performance and is superior to BMI as an indicator of scan reliability. A SLD>2.5cm was associated with higher IQRs, increased inter-operator discordance and lower overall accuracy. Care should therefore be taken when interpreting ARFI results in patients with a SLD>2.5cm.
